We have identified several consistent nonrandom chromosome abnormalities in human myelogenous leukemia, both chronic and acute. We have recently shown that a translocation between chromosomes 6 and 9 in acute leukemia is associated with an increase in marrow basophils. The breakpoint in chromosome 9 is in the same band as that in chronic myeloid leukemia, a disorder also associated with increased basophils. We found, however, that the cellular oncogene, c-abl remains on No. 9. Thus the breakpoint differs from that in chronic myeloid leukemia. In the inversion of No. 16, associated with acute myelomonocytic leukemia with abnormal eosinophils, we showed that the metallothionein gene cluster is split by the break in the long arm. By analogy with other rearrangements, we propose that the enhancer region of the metallothionein IIA gene acts as the switch to turn on an unidentified growth factor gene on the short arm of No. 16. We have shown the leukemic cells from patients who were previously treated with cytotoxic agents, such as radiation and/or chemotherapy for a primary malignant disease, had a very high frequency of loss of chromosome 5 and/or 7. Using leukemic cells with a deletion of part of No. 5 or No. 7, we have defined the critical chromosome region missing in each patient. We believe that these abnormalities may be markers of acute leukemia induced by mutagens, because they are rare in children and they are more common in the leukemic cells of patients who worked in industries where they were exposed to mutagenic substances than in patients not so exposed. There is a remarkably close correlation between the location of oncogenes and the breakpoints in translocations that occur in human leukemia and lymphoma. We are currently investigating the molecular nature of the breakpoint in the translocation involving chromosomes 8 and 21 in acute myeloblastic leukemia; our data show that an oncogene, c-mos, located in the same band as the breakpoint, remains on chromosome 8. In another translocation involving chromosomes 9 and 11, we showed that the interferon genes on No. 9 are split; the alpha chain remains on No. 9 and the beta chain moves to No. 11. An oncogene, c-its, moves from No. 11 to No. 9 adjacent to the interferon alpha chain genes. The identification of the genes that are involved and the resultant alteration in gene function will help in designing future therapy and in preventing these diseases. (K)